- Top of page
- Material and methods
Background: Eosinophilic airway inflammation is the hallmark of asthma, but it has also been reported in other conditions such as allergic rhinitis. We have tested whether the analysis of cells and chemicals in sputum can distinguish between patients with mild allergic asthma, those with allergic rhinitis, and healthy controls. The relationship between inflammation markers in sputum and nonspecific bronchial hyperresponsiveness to methacholine (BHR) (PD20 and maximal response plateau [MRP] values) was also evaluated.
Methods: We selected 31 mild asthmatics and 15 rhinitis patients sensitized to house-dust mite. As a control group, we studied 10 healthy subjects. Every subject underwent the methacholine bronchial provocation test (M-BPT) and sputum induction. Blood eosinophils and serum ECP levels were measured. Sputum cell differentials were assessed, and eosinophil cationic protein (ECP), tryptase, albumin, and interleukin (IL)-5 levels were measured in the entire sputum supernatant.
Results: Blood eosinophils and serum ECP levels were higher in asthma patients and rhinitis than in healthy controls, but no difference between asthma patients and rhinitis patients was found. Asthmatics had higher eosinophil counts and higher ECP and tryptase levels in sputum than rhinitis patients or control subjects. Sputum albumin levels were higher in asthmatics than in controls. Rhinitis patients exhibited higher sputum eosinophils than healthy controls. An association between sputum eosinophil numbers and MPR values (r=−0.57) was detected, and a trend toward correlation between sputum ECP levels and PD20 values (r=−0.47) was found in the rhinitis group, but not in asthmatics. No correlation between blood eosinophilic inflammation and lung functional indices was found.
Conclusions: Induced sputum is an accurate method to study bronchial inflammation, allowing one to distinguish between rhinitis patients and mildly asthmatic patients. The fact that no relationship was detected between sputum inflammation and BHR suggests that other factors, such as airway remodeling, may be at least partly responsible for BHR in asthma.
Prominent infiltration of the airway by activated eosinophils ( 1) and nonspecific bronchial hyperresponsiveness (BHR) ( 2) are constant findings of asthma. Several attempts have been made to demonstrate a causative relationship between both conditions. The results, however, have been contradictory, since some authors have reported a relationship between BHR and eosinophilic inflammation ( 3), but others have not detected such an association ( 1).
The mechanisms inducing a selective recruitment of eosinophils to the airways are not well known. There is increasing evidence that the TH2-cell subset, through the release of substances that selectively prime eosinophils, such as interleukin (IL)-5, or that enhance IgE-mediated responses (IL-4), plays a central role in the onset of bronchial inflammation ( 4). On the other hand, mast cells, which have been implicated in acute asthma, can also release a cytokine profile very similar to that described for TH2 cells ( 5). Mast cells can also produce other substances, such as tumor necrosis factor (TNF)-α or tryptase, that are involved in the chronic changes of the airway ( 5). Taken together, these data suggest that, mainly in the early stages of asthma, mast cells might play a role in the onset of airway inflammation ( 5).
Airway inflammation has been studied by indirect methods (peripheral blood) or invasive, fibrobronchoscopy-derived direct techniques. The induced sputum technique ( 6) and the treatment of the sample with cell-dispersion reagents ( 7) has proved to be a reproducible ( 6, 8–10) and valid ( 11–14) method to measure airway inflammation directly and noninvasively.
Although allergic rhinitis is probably a risk factor for asthma development, the relationship between the two disorders is not clear ( 15). A widely overlapping degree in sensitivity to methacholine has been reported between rhinitis patients and asthma patients ( 2). Therefore, it has been suggested that the identification of a maximal response plateau (MRP) might be a more discriminative index for differentiation of the two clinical conditions ( 16). Eosinophilic mucosal infiltration has also been reported in bronchial biopsies from allergic rhinitis patients ( 3, 17). Recently, Foresi et al. ( 18) have shown that induced sputum from seasonal allergic rhinitis patients exhibited an intermediate degree of eosinophilic inflammation, halfway between asthma and healthy controls, and that levels of sputum eosinophils in rhinitis patients were higher in those who also presented BHR ( 18). Furthermore, allergen inhalation by sensitized rhinitis patients could elicit an asthma-like eosinophilic response in bronchoalveolar lavage (BAL) ( 19). Taken together, these data suggest the lack of a clear border between allergic rhinitis and asthma.
The aim of our study was firstly to analyze the differences between cells and chemicals measured in induced sputum and blood from asthmatic patients and rhinitis patients who were sensitized to Dermatophagoides pteronyssinus, and from healthy subjects. Secondly, we aimed to assess whether inflammatory markers in sputum could distinguish between allergic asthma and rhinitis better than when measured in blood. Thirdly, we tried to determine whether sputum inflammatory markers were correlated to the baseline lung function and to the degree of BHR determined by the complete dose-response curves in which the MRP was examined.
- Top of page
- Material and methods
In this report, we studied a homogeneous group of D. pteronyssinus-allergic, untreated, and short-evolution (12–24 months) mildly asthmatic patients. We selected this type of patient, firstly, to avoid as far as possible, the influence of the bronchial remodeling process due to chronic airway inflammation on the results ( 23); and, secondly, to compare the capacity of induced sputum and peripheral blood to distinguish between patients with mild asthma and those with allergic rhinitis. Patients were grouped according to an exhaustive clinical history in which we confirmed that rhinitis patients had never experienced bronchial symptoms. In accordance with other reports, our results show higher eosinophil levels and ECP values in asthmatics than healthy subjects, both in blood ( 24, 25) and in induced sputum ( 8–10, 12, 26). Eosinophils and ECP values were also higher in sputum from asthma patients than rhinitis patients, but this difference was not observed in peripheral blood.
It has been reported that rhinitis patients exhibit a degree of eosinophilic inflammation both in bronchial biopsies ( 3, 17) and in BAL ( 27), but, to date, only two studies have compared eosinophilic inflammation in induced sputum from asthma patients and rhinitis patients ( 10, 18). In agreement with our results, these studies reported higher percentages of eosinophils and higher levels of ECP in sputum from asthma patients than rhinitis patients. However, when rhinitis subjects were compared to controls, Foresi et al. ( 18) reported higher eosinophil percentages and ECP levels in the first group, but Spanavello et al. ( 10) did not find such a difference. In our study, the sputum eosinophil percentage was higher in rhinitis patients than in healthy controls, but we did not find such a difference when we analyzed sputum ECP levels. Our results might be attributed to the dilutional effect that saliva exerts on fluid-phase chemicals when the entire sputum sample is analyzed ( 12). However, we think that the differences are most likely due to the features of the selected patients: all the rhinitis patients were sensitized to a perennial allergen, a fact that implies a continuous stimulus over the airway, and explains the higher eosinophil percentages found. The rhinitis patients had never experienced asthma symptoms, and although PD20 values were detectable in 50% of them, these values were much higher than the values reported by Foresi et al. ( 18). On the other hand, Spanavello et al. selected seasonal allergic rhinitis outside the exposure period, in which BHR was undetected ( 10). In keeping with their study, our study found that the eosinophil percentage and the ECP levels correlated in sputum from asthmatics, but not in sputum from rhinitis patients. These data strengthen the hypothesis that, although presenting airway eosinophilic infiltration, rhinitis patients have a lower degree of eosinophilic bronchial activation than asthmatic patients.
Although the former results indicate the existence of a relationship between airway eosinophilic activation and BHR, there is no consensus on this point. Some authors have found significant correlations among PD20 values and sputum eosinophils ( 8, 28) or sputum ECP levels ( 29, 30), whereas other studies have failed to find such a relationship ( 6, 10, 31, 32). The mechanisms determining the sensitivity (PD20) and the response (MRP) to the agonist are thought to be different ( 33). However, it is feasible that airway inflammation contributes, at least in part, to both ( 2): through the release of cationic proteins, eosinophils can damage the respiratory epithelium and increase the airway permeability, enhancing the sensitivity to the agonist.
On the other hand, the infiltration of the bronchial mucosa by inflammatory cells increases the thickness of the bronchial wall, causing a higher airway lumen stenosis for the same bronchial muscle contraction. ECP can also activate myofibroblasts ( 34), contributing to the characteristic remodeling process that causes the decrease of the lung elasticity forces ( 23). Accordingly, we found an association between sputum eosinophil counts and MRP values, and a trend toward correlation between sputum ECP levels and PD20 values in rhinitis patients. Such associations were not detected in the asthmatic group, perhaps because a degree of airway remodeling can be found even in mild and short-evolution asthma ( 1), and it would enhance BHR. Therefore, it is possible that the influence of airway inflammation in BHR is lower in asthma than in rhinitis.
IL-5 is the main cytokine priming eosinophilic chemotaxis and activation ( 4). The complete lack of detection of IL-5 in our study could be attributed to the dilution of the sputum sample or to a possible DTT-induced denaturalization of the IL-5. The latter contingency is not likely, since other authors have reported that sputum treatment with DTT does not affect IL-5 measurement ( 8). In accordance with our results, Louis et al. ( 30) have recently reported the absence of IL-5 in whole-sputum samples, despite strong spiking of the sputum supernatant with cytokines. Therefore, our results could be due not merely to a dilutional effect, but to the binding of cytokines to mucus, an effect that would make detection of IL-5 in the supernatant difficult after filtering of the sample ( 30).
To evaluate the role of mast cells in the airway inflammation, we measured the levels of tryptase, their specific activation marker. In agreement with other reports ( 8, 9), sputum tryptase detection was low but selective for asthmatic patients. The degree of correlation detected among tryptase, ECP, and albumin in sputum suggested simultaneous activation of eosinophils and mast cells. Tryptase, through the enhancement of smooth-muscle contraction, the activation of collagenase, and the disruption of respiratory epithelium intercellular unions ( 5), seems to be deeply involved in the bronchial muscle response to agonists, in the airway remodeling process, and in the fragility of the respiratory epithelium. Therefore, we would have expected to find a relationship between levels of tryptase and the MRP values. However, we were unable to detect such an association, a fact that might be attributed to the low number of asthma patients in whom MRP was identified.
In agreement with other reports ( 8, 9), we found the levels of albumin in sputum from asthma patients to be higher than in healthy controls. When the rhinitis group was analyzed, albumin levels did not differ from those of asthmatics or control subjects. Sputum albumin levels were directly correlated with sputum ECP levels in both groups and with sputum tryptase levels in the asthmatic group, suggesting that activated eosinophils and mast cells, through the release of vasoactive mediators, increase vascular permeability.
In conclusion, we have demonstrated that analysis of the eosinophilic inflammation in sputum samples allows one to distinguish between patients with mild asthma, those with allergic rhinitis, and healthy controls. Perennial allergic rhinitis patients present a degree of eosinophilic inflammation in sputum intermediate between asthma patients and control subjects. The absence of a closer correlation between sputum inflammation and BHR might be due to the multifactorial origin of BHR. Analysis of induced sputum samples reflects the inflammatory changes occurring on the intrapulmonary epithelial lining fluid but gives no information about the chronic changes of the remodeling process that have been established at the airways.